Why are domains not formed in paramagnetic materials? Why are domains formed in ferromagnetic materials, but not in paramagnetic ones?
 A: Domains do form in paramagnetic materials. The distinguishing feature of a ferromagnet is that one single domain grows extensively as the system gets larger. But for a finite-size system near the (corresponding infinite system's) phase transition, the two phases behave identically if you only look at small local clusters of spins.
A: The obvious answer to the question “Why are domains formed in ferromagnetic materials, but not in paramagnetic ones?” is because if domains were formed in a paramagnetic material then it would be ferromagnetic. 
I suppose that the real question you have is “What is the difference between ferromagnetic materials and paramagnetic materials that causes formation of domains in one and not in the other?” 
Both paramagnetic and ferromagnetic materials have unpaired spins that give individual atoms a magnetic moment. The difference is due to the collective behavior. In paramagnetic materials the atoms align innreaponse to an external magnetic field, but in ferromagnetic materials the alignment is spontaneous. 
There are two competing effects. Roughly speaking when the unpaired spins of neighboring atoms are aligned then the Pauli exclusion principle keeps them further away from each other which reduces their electrostatic energy. On the other hand when the unpaired spins of neighboring atoms are antiparallel their magnetic energy is reduced. 
The behavior then is based on which competing effect is stronger. For ferromagnetic materials the first effect is stronger. The domains are caused as the atoms go to the lowest energy state which has aligned magnetic moments to reduce the electrostatic energy. 
